We often use ride-sharing services in congested areas and have envisioned ways to improve the experience when supply and demand for drivers and rides are not in parity. There also exist other efficiency enhancement opportunities when location conditions are not ideal for pickup/dropoff spots. There is improvement potential in areas where surge pricing results in a cancelled ride/negative review/non-booking, when a user has run out of battery on their primary smartphone after an eventful evening, wants to be recognized by the driver easier, and/or wants to flag down a driver passing by.

What it does

Inspired by our team name, Cobra Kai, Miyagi was developed as a service to increase efficiency in high volume areas by providing static and dynamic virtual transit hubs for customers and drivers to expedite movement in an otherwise unmanaged way. This enables the possibilities to provide drivers with a more streamlined/steady customer flow, provide ride-share companies data on usage patterns, and provide users an incentive to use this service. Through the use of our app, drivers will know in advance where "the party" is taking place and when it will be letting out, giving them the opportunity to congregate in the general location of our hubs, and their accompanying zones, aka "Kreeses" to speed up pick up. We also hope to be able to offer incentives to riders to opt into a "Kreese to Kreese" trip at a decreased price. This also benefits the drivers, as "Kreeses" are places that, either historically or dynamically, have a high volume of riders, drastically decreasing their downtime and alleviating the need to hunt for riders. We can also use them to spread out riders so there isn't just a big clump of people in one spot. Uber is attempting to do this right now with their suggested pick up spots, but since we will know where the riders and drivers are located we can direct everyone to a better location to get in and out.

How we built it

The first deliverable of the Miyagi service is an Android Application for the user and driver powered by a node.js backend. The first screen in the Miyagi app is a map similar to other ride sharing service providers. What makes us special is our ability to sort, queue, and match up drivers and riders.

Challenges we ran into

We faced many challenges along the way. This included but is not limited to:

  1. Unstable internets -> fixed this by creating a hotspot with a raspberry pi (The DoJo)
  2. PrimeTime/Surge pricing indication of dead spots, hot spots, or supply/demand imbalance and using this data to produce dynamic hubs, and using traffic data to lay out Kreeses
  3. Getting real driver location to produce better dynamic hubs
  4. Determining static hub locations
  5. Designing a system which queues drivers and riders

Accomplishments that we're proud of

Able to get a working prototype end to end Completed market research with Lyft and Uber drivers Preliminary data analytics to drive decisions

What we learned

  1. We spoke with many Lyft and Uber drivers about pain points that they experience and ways that their process can be improved
  2. We learned a LOT about surge/primetime pricing and how they are calculated and placed on the map
  3. We learned about how to put together a presentation to demonstrate our fledgling startup.

What's next for Miyagi

Ad-hoc Virtual Flag down of rideshares Partnerships with local vendors/bars/brick and mortar establishments to order an uber


If supply of drivers does not exists

Share this project: